To unlock the full potential of voice as information, you need to look to the much ignored call center world. I say this with a twinge of reluctance, but call centers have been at the leading edge of getting the most of voice for over a decade. The rest of the service provider enterprise world needs to get onboard as those sectors fully embrace voice over IP and an all IP world.

When you call a contact center outside of the U.S., you may be asked to use your voice to identify yourself. Voice biometrics -- the whole "my voice is my password" idea -- has been integrated into call center operations from Turkey to New Zealand. Use of a pre-registered voice print has cut an average of 28 seconds off the average length of call to New Zealand government contact centers  -- multiply that out by the number of agents times the number of calls per day to get some significant savings.

Everyone has heard the "Your call may be recorded" notice -- let's not kid around. Every call is recorded. Call centers treat every call as a source of information that can be data mined -- "Big Data" if you prefer-- for information. 

There are three examples applicable to every business where call recording is a vital tool. First, call recording can be used as documentation to be referenced in the case of disputes or problem resolution.   If a customer believes he was told something was wrong on a previous call, another agent or supervisor can reference either transcripts of previous calls and/or listen to previous calls to determine what transpired.

Improvement in business processes is the second area where call recording can play a role. By reviewing the "best performers" by such metrics as most sales closed or most customer service calls effectively resolved, best practices can be identified, documented, and communicated to the rest of the specific business unit. From a management standpoint, the purpose for call recording should be to try to coach "up" the rest of the team to be better performers.

Voice analytics is voice as big data. A large volume of recorded calls are translated from speech into text. Once done, the masses of transcripts are sifted through for key words and phrases to identify best practices, successful marketing offers, and competitive intelligence. For instance, a service provider could look for the words "cancel" and "offer" if they see a sudden spike in customer churn, then drill down to find out who is offering what to steal away customers.

My reluctance in holding out call centers as the prime example of SmartVoice is two-fold. You may remember the push about a decade ago of how call center software would turn everyone in the company into a customer service agent, enabling subject matter experts to be reached wherever they were in the world. It didn't work out because companies wanted their workers outside of the call center to focus their specific assigned tasks, rather than dragged into conversations at any time day or night. 

Providers of SmartVoice class services are going to have to adopt thinking and business models to appeal to a broader audience. Vendors are used to receiving top dollar from a limited number of deep-pocketed customers. Evolving services to service a general corporate VoIP infrastructure and as a cloud offering to the SMB community may prove to be difficult for many.